Monday, September 11, 2017

Socialism Reconsidered Part 5 - Wage Subsidies and Confounding Expectations

This is part 5 in an occasional series.  In this piece we grapple with the economics of income inequality as it applies to our national politics and make a case for doing something real and substantial about it that is highly visible, well before the coming election in November 2018.

* * * * *

We live in a time where in the public consciousness an emotional appeal can win over a rational argument.  Maybe that has always been true.  I don't know.  It is clearly evident now, particularly on why the base doesn't abandon Trump and instead ignores the matter of Russian involvement in influencing the election and the Trump campaign's role in that, all the while staying loyal to their guy.  Among the two parties, it seems the Republicans in general, and this White House in particular, are far better at making the emotional appeal.  The social science suggests that those who are disposed to an authoritarian view are especially welcoming of such an appeal.  Further, it seems the Democrats have pulled the short straw on which party is to be trusted.  And, of course, the Democrats are the minority party now.  Even if they make an excellent rational argument, for the time being the argument is talk only.  We all know that talk is cheap.  Under these conditions does the message get
through?  Does it get a full hearing?  Is it then believed?

On the flip side, money talks in a way that people listen.  The core idea in this essay is to use money, preferably quite a lot of it, in a way to deliver the message that the Democrats want to deliver, but not through the normal political channels, such as spending the money on TV ads.  Instead, spend the money on the people whom the Democrats claim they want to benefit, ordinary working people.  This would make the message credible.  It would also make the message quite different, a game changer if you will.  Nobody has done this before in a political campaign.

One other point to make here is that while individual candidates need to make themselves known to the voters, this time around there is no reason to run attack ads against the President or against the Republican Party.  The track record speaks for itself.  Rubbing it in won't help.  And we really don't need any more negative messages.  We need something constructive, illumination on a way to move forward.  That's what the campaign should be about.

The money that is raised would be generated by donations, mainly from upscale voters.  In appealing to them for donations two distinct points need to be made.  First, their donations signify that they are willing to have their taxes raised, after the election in 2020 if not the election in 2018, for the good of the order.  Conventional politics says that talk of raising taxes is a loser.  So politicians tend to talk about the beneficiaries of government spending, but then mumble about how to pay for it.   This won't work now.  An aggressive case needs to be made to the effect: (1) the current system doesn't work because it screws the little guy, (2) for the system to work there needs to be substantial income redistribution toward the little guy, (3) well off people must bear their fair share of the burden; expecting the uber rich to pay for it all is unrealistic, and (4) the previous point must be cast in ethical terms as a matter of social responsibility; people need to provide service to their country in ways that matter, not just by serving in the military; now the need is for income transfers.  The debacle about repeal of the Affordable Care Act should make all of this abundantly clear.

As near as I can tell, there has been a lot of talk about (1) and (2) recently, but hardly any talk at all about (3) and (4).   The second point about appealing for donations has already been made.  Much of these donations would go directly to ordinary working people, who are barely getting by now.  During the campaign, these income transfers would be for demonstration purposes of what will be possible, should the Democrats retake the majority in Congress and ultimately retake the White House as well.  In addition, there would be an experimental aspect to such a program, pilot projects if you will.  So they would be studied for their effectiveness, with the possibility of recalibrating and that leading to future redesign.  While the need for income redistribution is evident, how it should be executed is far less obvious.  One should anticipate that some experience in the execution can help make the process more effective.  Thus, we should use the phase where income transfers are purely voluntary during the campaign to inform the design when income transfers become law.

This suggests several possible points of failure.  I will mention a few that occur to me now.  (a) Insufficient donations are generated because potential donors are not convinced about the benefits from the program.  Indeed, potential donors may be driven to vote Republican for fear that otherwise their taxes would be raised.  (b) Donations don't reach their intended recipients.  The funds are pilfered en route.  (c) While the recipients do benefit from the transfers the effect appears mild and hence doesn't offer a compelling story on which to base subsequent donations.

To this list one might add a hybrid between (b) and (c), namely that the recipients have debt for which they have been somewhat delinquent in paying off.  The creditors who hold that debt swoop in to collect the money for themselves, in which case the bulk of the transfers go to the creditors rather than to the intended recipients.  Strictly speaking, this wouldn't be pilfering.  Indeed maybe a bit of this would be a good thing, getting the recipients to reduce their debt overhang.  But this should be done in a balanced way and there would be a need to ensure that balance as an outcome.

A plan that implemented a voluntary income transfer program would need to address these various points of failure to make them far less likely.  Here are a few preliminary thoughts on that.

Among the points of failure (a) will be the hardest to overcome.  I think it is useful to consider donation for the purpose of income transfer as an innovation, so that we can apply the language and methodology of diffusion of innovations to the problem.  A first step would be to identify innovators and early adopters.  People who do fundraising as their business may already have a list of such folks in mind.  In turn, these early givers can then be employed as exemplars to encourage others to follow in their footsteps.  While I have emphasized upscale voters in the above discussion, it would help if a few high rollers were among the early group, so sufficient funds were generated that some projects could get underway quickly.

Making the project visible will help to eliminate fraud as listed in (b), but of course there needs to be sufficient monitoring as well.  There will need to be accountants who track the flow of funds as well as a local coordinator for each project to ensure things are going as intended.  Recipients must then report in (though how that should be done needs to be determined) both on their receipt of the funds and on broad strokes uses of the money.

The recipients within a single project should be geographically concentrated as a partial way to address (c).  They should be working at the same places and living in the same communities.  This will enable an outside observer to see whether there are productivity effects from the transfers at the places of work and whether there are multiplier effects within the community that produce an uptick in economic activity.  The transfers must also be of sufficient duration, say at least a year, for behavior to adjust to them.  Temporary, short lived transfers should not be expected to produce much effect.

This latter observation suggests that the trade-off between the number of projects, the scale of the projects, and their duration should be biased toward having fewer projects so that one can get enough oomph from those projects that are undertaken.  Of course, this must be tempered by the available revenues needed to make a project a reality.

* * * * *

In this section I want to focus on a specific type of income transfer, a wage subsidy that for the recipients is meant to be a proxy for an increase in the minimum wage.  Before getting to the mechanics of the subsidy, here are some caveats to consider.

Invariably in considering income transfers to reduce income inequality, the question will come up as to whether the recipients are worthy of receiving the transfer.  In turn, because we've already considered this above, worthiness of the recipient will matter for whether donations are made.  So here we will focus on those who are already working, who deserve to be making a living wage.

What then of those who are no longer counted in the labor force participation measures, as they've become too discouraged to look for work?  They need income too, no doubt.  The view here is that to address their needs broadly a program of voluntary income transfers is inadequate.  However, some of these people might become encouraged to look for work, were the market wage substantially higher.  Indeed,  others who are currently working elsewhere and those unemployed who are still actively looking for a job will be attracted by work at higher wages.  So the program needs to be able to expand beyond current employees, as long as employers who participate in the program are willing to hire them.  As it will be for profit businesses that employ these people, hiring additional employees will only make sense if there is a business case for doing that.  This is precisely the multiplier effect mentioned above.

Next, small business will be targeted as the likely targets of income transfers for their low wage employees.  There are a few reasons for this.  A dramatic increase in the minimum wage might constitute a substantial burden on small businesses, which don't have retained earnings nor sufficient cash flow to finance the increase.  So the subsidies may point to a permanent policy targeted at small businesses to help them absorb a substantial increase in the minimum wage.  Small businesses are known as job creators.  If we can increase not just employment but also well paying jobs, that would be a great accomplishment.  Then too, large businesses that are philosophically opposed to an increase in the minimum wage might push back at the voluntary program, aiming to disrupt it.  Better to not approach them at all, especially at first.  This constitutes a "judo approach" to diffusion of the idea.  Win the battles you can win and fight those first.

It may be, however, that some large companies which employ a substantial number of low wage workers and which have sufficient cash reserves to pay those workers more should they decide to do so, take note of the voluntary income transfer program, its visibility and its effectiveness.  Then, some of these large companies might opt to replicate the pay schedule under the program within their own companies, both because of the productivity impact on employees and because of the goodwill generated with customers as a result.  Were this to happen, it should be welcomed as an encouraging development.  The goal is not to make the program as big as possible, although it may seem that way to those who are wrapped up in running it.  The goal is to make as many low wage workers as possible earn a living wage.  Imitation of the program that is done privately would clearly help to achieve that goal. Further, it would be an indicator that the program is working.

There is one more caveat, this time a political one.  Ideally, the projects are diverse, both geographically and in the nature of the communities they serve.  It would be far better to have one project in each Congressional district than to have a cluster of projects in a few Congressional districts and no project whatsoever in many other Congressional districts.  There need to be urban projects, suburban projects, and rural projects.  There need to be projects in red states, blue states, and purple states.  Some projects may benefit one gender over another, by the nature of who does the subsidized work in those communities.  There then needs to be other projects where the benefit goes the other way.  This is likewise true for projects that benefit people of a certain race or national origin.    Once income transfer via an increase in the minimum wage becomes law, the impact will be felt across the country.  For people to vote to support that outcome, they need to believe it will actually happen.  The diversity of projects is a necessary precondition to encourage that belief.

Let us turn to description of the subsidy policy. Eligible employees will receive a subsidy based on their current hourly wage.  The subsidy added to the hourly wage will then be at least $15/hour, the minimum wage in the Democrats proposal.  (The current Federal minimum wage is $7.25/hour.  Many states have a minimum wage that is in excess of the Federal minimum wage.)

Let's note the following issue, which impacts how the subsidy policy is constructed.  Consider two different people who currently earn less than $15/hour but who are paid at different wage rates.  Should the differential between their current wages matter in determining what they earn after the subsidy policy is introduced?

Much of the public discussion about raising the minimum wage ignores this issue.  But, as I'm trained in microeconomics, I have a healthy respect for wage differentials.  They exist for some reason.  Here are a handful of possible explanations for them.

a)  The higher paid person is receiving a reward for seniority.  That person has been on the job longer.  The business makes it a policy to reward its long-time employees.
b)  The higher paid person does a different job that entails more responsibility and is being rewarded for that.
c)  The higher paid person does a job that is more unpleasant (such as work the night shift).  This sort of wage differential is called a compensating differential.
d)  The higher paid person works for the employer across the street, who is more generous paying her employees than the other employer, because she feels she gets more out of her employees that way.

While this is a pretty good list, there may very still be other good reasons to have wage differentials that we would deem good and productive, rather than the result of favoritism or discrimination.  So it is worth contemplating the trade-off between attaining a living wage and maintaining wage differentials across jobs after the subsidies are implemented.  To illustrate the various possibilities consider the three policies illustrated in this Excel spreadsheet.  (The preview may be sufficient to see what is going on.  Alternatively, the workbook is freely available for download.)

Each policy occupies three columns - a wage before the subsidy, the subsidy itself, and total compensation afterwards.  The first policy is called Subsidy with Total Wage Compression.  In this policy the subsidy plus the prior wage adds to $15/hour for all employees who had been earning less  than $15/hour. This policy exactly imitates in its consequence what raising the minimum wage to $15/hour would do.  For our purposes, we can think of it as one extreme.  The other extreme is given by the third policy, which is called Subsidy with No Wage Compression.  In that policy the subsidy is a fixed amount, regardless of the pre-subsidy wage.  As a consequence, the prior wage differentials are maintained in full.  Further, the sum of the pre-subsidy wage and the subsidy may well exceed $15 and there is no maximum pre-subsidy wage above which the subsidy would not be paid.   Thus, while the third policy does well at maintaining the prior wage differentials, it supports high wage earners as well as low wage earners, although the former are already earning a living wage.

The second policy, called Subsidy with Partial Wage Compression, is a hybrid of the two extreme policies.  Personally, I favor it over the other extremes as it represents a compromise between the competing objectives: getting recipients to earn a living wage, encouraging donors to want to give for this purpose, and maintaining prior wage differentials.  It should be noted that one can have other hybrid policies, each that is closer to one of the extremes.  There is nothing here to say which of these hybrids would be best.  The one that is illustrated has been chosen for simplicity of illustration only.   It represents the midpoint between the two extremes.

For any given project, the wage distribution among the employees who will receive the subsidies should matter about which policy is preferred.  If most of the employees have prior wages concentrated at the lower endpoint of the distribution, a no wage compression approach will be approximately optimal, as there will be very few subsidy payments to high wage earners.  Alternatively, if there are more employees who have been earning well above the lower endpoint, then having more wage compression in the policy will be preferred, so as not to overly reward employees who were already paid reasonably well in advance of the subsidy. 

In announcing the program, however, there needs to be one policy in place that applies to all the projects at once.  That would be the fair thing to do.  Before the program is launched, then, the appropriate policy would need to be determined.  The above discussion is meant only to consider the factors that would go into such a determination.  While I would expect some hybrid policy to emerge, the particular policy must be decided by the decision makers within the program.

Let us turn to providing cost guesstimates for the projects and then considering project scale as a consequence of the costing exercise.  A few calculations are provided on the second worksheet of the Excel workbook linked above.  To begin, an assumption is made that the average subsidy per hour is $7.50.  This is below the maximum possible subsidy, $7.75.  But it is still a substantial subsidy.  So it can be thought of as generating an upper bound on cost.  Then, it is assumed that employees work full time, a 40 hour week.  It may be now that many employees work part time, so they can juggle work and school or work and some other family obligation.  If this is done by adult employees, that juggling should be accommodated.  If that sort of thing is commonplace, it means more employees can be part of the project and that when we discuss number of employees, we are talking about full-time equivalents.  The thought here is that kids in high school who are working a part time job should not receive these subsidies, even though were the minimum wage raised to $15/hour they'd then get paid at that rate.  This program is aimed at adults who need to make a living wage.  That should be the focus. 

It is then assumed that the employees work 50 weeks a year, to come to an annual cost calculation.  Round numbers are chosen to make the calculations easier to follow.  The goal here is not precision in the costing.  Rather it is to develop a method for determining costs when more realistic numbers can be supplied.  Here we just want the cost guesstimates to be in the ballpark.

Not in the spreadsheet, but an important assumption that underlies this exercise, is that projects would start with between 250 and 500 employees (or full time equivalents) and then be able to grow to between 500 and 1,000 employees.  Until that maximum is reached, additional subsidy funds would be forthcoming as the project grows, with the additional funds there to enable that growth.  Once the maximum is reached the total subsidy the project receives would be frozen.  Further growth in subsidy would need to be locally financed.  This limit, though arbitrary, is there so one project doesn't hog too many resources and, as a consequence, to better allow other projects to be started.

Given this assumption and the prior assumptions, if the average sized project has 500 employees, one can compute the direct annual expenditure on subsidy for such a project.  It is $7.5 million.  A maximal sized project would entail twice that expenditure on subsidy.  The last bit on coming up with full project cost is to get a handle on overhead/administrative costs for the program.  I have very little sense of what is realistic here, other than to note that many of the overhead costs will be fixed costs, so as a fraction of overall costs they will decline as the program gets bigger.  But to keep the calculations simple I suggested a 20% rate to compute overhead.  (Again, that makes the calculations simple.)  With that assumption, the full cost of an average sized project (subsidy plus overhead) would be $9 million and for a maximal sized project it would be $18 million.

These numbers can be used as a first pass at how much revenue needs to be generated from donations to achieve certain targets - say 100 projects in total.  And for the a higher target - getting one project per Congressional district - then to paraphrase Everett Dirksen, now you're talking real money.  With the same sort of calculations one can also talk about impact.  On the order of 200 projects would produce 100,000 recipients of subsidy.  Surely, a program of that magnitude would generate substantial visibility and, we hope, derivative impact about wanting to make program outcomes permanent.

The last bit to consider in this section is how projects would be selected.  As a full process would have to be negotiated by those running the program,  here I will contain myself to talking about fairness and elements to help assure that that process is perceived as fair.  Donors should be enlisted to support the program but need to be excluded from project selection, as they might otherwise be expected to pick favorites and that would undermine fairness from the get go.  Early projects need to be selected with an eye toward generating interest and excitement in the program.  But subsequent projects need to conform with the diversity needs that the program requires.  In other settings, such as college admissions, applying diversity criteria can create some backlash among applicants who perceive they are being treated unfairly.  There is no magic elixir to apply that would preclude such perceptions.  The best that can be done is to heavily promote the diversity criteria ahead of time, at the inauguration of the program.  Consistency is needed in applying those criteria as various candidate projects compete with one another for funding.

This issue of fairness doesn't just apply to the projects themselves.  As Thomas Edsall's latest column indicates, The Struggle between Clinton and Sanders Is Not Over.  This program should not favor one side over the other in that struggle.  That is a tall order.  At a minimum, it means that the board which engages in project selection must have representatives from both sides.  (It must also have representatives from donor groups and experts in community development, to have the right balance.)   Further, in preliminary discussions before the program is operational, each camp should be solicited about how fairness might be attained and whether, given the programmatic goals, infighting can be resisted.  It seems evident that such infighting would become public and then undermine the objectives of the program.  This is not to say that there can't be heated discussion during the formative period where the program is being developed.  People need to get their issues on the table and, as best as possible, those need to be addressed.  But that needs to happen early on.  In the ideal those can be resolved up front.  Once the program is underway, second guessing the process would be unhelpful.  The participants need to understand that and agree to mute subsequent objections after the program is well underway.

One last point about fairness is that the duration of the program matters.  Since the program is tied toward electoral ends, if those are achieved in full then by the end of 2021 the program would be abandoned, as legislation on the minimum wage would have been implemented by then in accord with ideas suggested in this proposal.  One might then anticipate that the likelihood of achieving program goals would matter for fairness as perceived by donors, potential projects, and project participants.  This means that fundamentally ethical matters will be blended together with pure expediency, normally not something that commends itself.  So be it.  I see no other way for this to happen.

* * * * *

I want to wrap things up.  So here I'll talk a little about the motivation for writing this piece.  Of course, I am disturbed by the current White House, the Republican domination of Congress, as well as their domination of the vast majority of State Houses.  Their anti-tax anti-government ideas are wrong headed, in my view.   So, on the one hand, the thought is simply to propose an alternative that would have popular appeal, precisely because it did help the little guy, and in the process change the electoral calculus.

Beyond that there is a concern that the internal politics within the Democratic party, as detailed in the Edsall piece linked above, might derail this goal, in spite of the incredible unpopularity of the current administration.  I'd like to see movement toward a more consensus view, away from the factionalism that is evident right now.  The thought was that to achieve consensus one needed to generate some synthesis where both sides matter in the product that is ultimately produced.   I am a naïve outsider to this internecine conflict.  I did vote for Hillary Clinton in 2016, but I am quite sympathetic to income redistribution politics away from the well to do and toward working class people. I really don't know how many people in the professional class (people in top 20% but below the top 1% in the income distribution) have similar beliefs, but my hope is that many do.  If so, my appeal should resonate with others.

The other point that motivates me here, sad to say, is that I haven't seen others who are similarly motivated as I am, described in the previous paragraph, come up with their own solution to reach a consensus view.  In general, Democrats like to duke it out and let the best participant win.  Normally, that is not a bad position to have.  But now, if infighting by the Democrats limits their electoral success in 2018 and 2020, that would be a disaster, worse than Harvey and Irma combined.

So, at a minimum, even if people otherwise find issue with the ideas advanced here, my hope is that this piece will encourage people to think of what synthesis might be identified, so a credible rapprochement can be found within the Democrats and so faith can be restored that we will stop shooting ourselves in the foot in our politics and make progress possible thereafter.

No comments: